{"id":15175,"date":"2022-04-20T03:53:28","date_gmt":"2022-04-20T03:53:28","guid":{"rendered":"https:\/\/blog.selectstar.ai\/kr\/?p=15175"},"modified":"2022-04-20T03:53:28","modified_gmt":"2022-04-20T03:53:28","slug":"klue","status":"publish","type":"post","link":"https:\/\/dev.selectstar.ai\/ko\/klue\/","title":{"rendered":"Open Datasets &#8211; KLUE"},"content":{"rendered":"<p>[vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div class=\"pix-content-box card      custom-responsive-4373055   rounded-lg bg- w-100  \"   ><div class=\"\" style=\"z-index:30;position:relative;\">[vc_column_text]<\/p>\n<p style=\"text-align: left;\"><span style=\"font-size: 14pt;\"><strong>\ud83d\udd11<\/strong> <strong>In 10 minutes you will learn:<\/strong><\/span><\/p>\n<ul class=\"p-rich_text_list p-rich_text_list__bullet\" data-stringify-type=\"unordered-list\" data-indent=\"0\" data-border=\"0\">\n<li data-stringify-indent=\"0\" data-stringify-border=\"0\">The definition and significance of KLUE Benchmark<\/li>\n<li data-stringify-indent=\"0\" data-stringify-border=\"0\">The types of Korean language datasets KLUE consist of<\/li>\n<li data-stringify-indent=\"0\" data-stringify-border=\"0\">How accurate Datumo\u2019s NLI dataset is, compared to SNLI(Stanford Natural Language Inference) and MNLI(Multi-Genre Natural Language Inference)<\/li>\n<\/ul>\n<p>[\/vc_column_text]<\/div><\/div><div id=\"el1646799961152-e3ee06c0-4e82\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_row_inner][vc_column_inner width=&#8221;1\/5&#8243;]<div class=\"d-inline-block2 d-flex flex-column flex-sm-row w-100 text-center2 align-items-center2 justify-content-center justify-content-sm-start  \"><div class=\"pix-md-circles d-inline-flex align-items-center align-middle justify-content-center justify-content-sm-start\"><\/div><\/div>[\/vc_column_inner][vc_column_inner width=&#8221;4\/5&#8243;]<div id=\"el1650448025755-3da2e3e5-e893\" class=\"w-100 d-block \"><\/div>[vc_column_text]<\/p>\n<p class=\"indexstyle__SubTitle-sc-eckm8t-5 JOTtw\">Korean Language Understanding\u00a0Evaluation Benchmark<\/p>\n<p>[\/vc_column_text]<div id=\"el1650448001264-6722a641-4e7a\" class=\"w-100 d-block \"><\/div>[\/vc_column_inner][\/vc_row_inner]<h6 class=\"pix-badge-element d-inline-block mr-1 \"  ><span class=\"badge font-weight-bold secondary-font bg-primary-light  \" style=\" \"><span class=\"text-primary\" style=\"\">Korean<\/span><\/span><\/h6><h6 class=\"pix-badge-element d-inline-block mr-1 \"  ><span class=\"badge font-weight-bold secondary-font bg-primary-light  \" style=\" \"><span class=\"text-primary\" style=\"\">Natural language<\/span><\/span><\/h6><h6 class=\"pix-badge-element d-inline-block mr-1 \"  ><span class=\"badge font-weight-bold secondary-font bg-primary-light  \" style=\" \"><span class=\"text-primary\" style=\"\">NLP<\/span><\/span><\/h6><div id=\"el1646121660539-c86f72c7-0277\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h1 class=\"text-heading-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Considers unique characteristics of the Korean language<\/h1><\/div><\/div><\/div><div id=\"el1650294698986-a1b962b5-ef42\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653300469410{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;]<\/p>\n<div class=\"pix-el-text w-100 text-left \">\n<p>The KLUE(Korean Language Understanding\u00a0Evaluation Benchmark) paper, written using Datumo\u2019s datasets, has been accepted in NeurIPS 2021, a world-renowned AI conference.<\/p>\n<p>Natural language processing(NLP) has been an on-going research throughout the world. However, there were limitations in utilizing open datasets as the basis for Korean NLP research because most of them were in english, which made it difficult to produce precise results taking the unique characteristics of the Korean language into consideration.<\/p><\/div>\n<div class=\"d-inline-block \"><\/div>\n<p><img fetchpriority=\"high\" decoding=\"async\" class=\"wp-image-15344 aligncenter\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/\ud074\ub8e8\ub85c\uace02-1024x181.png\" alt=\"\" width=\"870\" height=\"154\" \/><\/p>\n<p>In order to solve this problem, the startup Upstage held hands with 10 other institutes including Korea Advanced Institute of Science and Technology(KAIST), New York University(NYU), Naver, and Google to build KLUE benchmark.[\/vc_column_text]<div id=\"el1650431947080-bf37527b-f26a\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650847594010-704deed1-7839\" class=\"w-100 d-block \"><\/div>[vc_row_inner][vc_column_inner width=&#8221;1\/6&#8243;]<div class=\"d-inline-block2 d-flex flex-column flex-sm-row w-100 text-center2 align-items-center2 justify-content-center justify-content-sm-start  \"><div class=\"pix-lg-circles d-inline-flex align-items-center align-middle justify-content-center justify-content-sm-start\"><\/div><\/div>[\/vc_column_inner][vc_column_inner width=&#8221;5\/6&#8243;]<div id=\"el1650847594011-02537b23-96d0\" class=\"w-100 d-block \"><\/div><div class=\"pix-el-text w-100  \" ><p class=\"  text-body-default  \" >SungJoon Park, Upstage AI research engineer &amp; Chief project manager of Project KLUE<\/p><\/div>[vc_column_text]<\/p>\n<p style=\"text-align: left;\"><span class=\"notion-enable-hover\" data-token-index=\"0\" data-reactroot=\"\">\u201cWhile building the KLUE dataset with Datumo, we were most impressed by their data quality assurance system. Despite the intricacy of the data and the tight deadline, Datumo was able to provide specific guidelines for the workers to guarantee data consistency. They also made sure to train and select qualified workers, and inspect the entire dataset. We believe that KLUE, the representative Korean NLP benchmark dataset, was able to come into the world, owing to Datumo\u2019s capability and passion.\u201d<\/span><\/p>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650442416450-43ff98a7-814a\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653300719216{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;]Amongst the eight Korean natural language understanding (NLU) tasks, the ones Datumo took care of were as follows:<\/p>\n<ul>\n<li>Topic Classification<\/li>\n<li>Semantic Textual Similarity<\/li>\n<li>Natural Language Inference<\/li>\n<li>Machine Reading Comprehension.<\/li>\n<\/ul>\n<p>A number of crowd-workers from Datumo\u2019s crowd-sourcing platform &lt;Cash Mission&gt; swiftly collected and annotated KLUE datasets. We\u2019ll take a further look at the four tasks.[\/vc_column_text]<div id=\"el1650442319658-303ff8d9-2f7f\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h3 class=\"text-body-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Topic Classification, TC<\/h3><\/div><\/div><\/div><div id=\"el1650294801245-81493317-8694\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653300764171{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;]As shown in the photos below, this task required crowd-workers to categorize news headlines into various topics such as politics, economics, society, IT, and more. The categorization does not simply rely on the headline including particular keywords, but on whether the content of the headline is related to certain topic or not.<\/p>\n<p>&nbsp;<\/p>\n<p><img decoding=\"async\" class=\"aligncenter wp-image-15211\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-2.07.40-PM-1024x457.png\" alt=\"\" width=\"720\" height=\"321\" \/><\/p>\n<p>&nbsp;<\/p>\n<p>Once three crowd-workers chose maximum of three topics for each news headline, the headline was categorized as the most-voted topic. In order to maintain data accuracy, workers were also requested to report any headline that includes personally identifiable information (PII), expresses social bias, or is hate speech. The reported headlines were discarded after manual review.[\/vc_column_text]<div id=\"el1650436966864-c646ca60-7eb9\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h3 class=\"text-body-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Semantic Textual Similarity, STS<\/h3><\/div><\/div><\/div><div id=\"el1650442668981-555f8d8b-1571\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653300795255{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;]Semantic textual similarity (STS) is necessary to measure the degree of semantic equivalence between two sentences. For this particular task, it was essential to make sure that it is not about monitoring if two sentences had the same exact words used, but about making sure the two sentenced had the same meaning.<\/p>\n<p>&nbsp;<\/p>\n<p><img decoding=\"async\" class=\"aligncenter size-full wp-image-15215\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-2.12.49-PM.png\" alt=\"\" width=\"261\" height=\"457\" \/><\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<p>\u201cThe one that washes well\u201d and \u201cwhich cleans clothes better\u201d do not have any common vocabulary used, but we know that both mean the same. As such, apart from using the same exact vocabularies, we make sure to compare the speakers\u2019 purpose, nuance, and overall emotion to train AI and make the model smarter.[\/vc_column_text]<div id=\"el1650847791055-2a7fe6e7-07d6\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h3 class=\"text-body-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Natural Language Inference, NLI<\/h3><\/div><\/div><\/div><div id=\"el1650436978481-aa98d546-b3a1\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653459817005{margin-right: 7% !important;margin-left: 7% !important;padding-top: 40px !important;padding-bottom: px !important;}&#8221;]The purpose of NLI is to train AI models to infer the relationship between a hypothetical sentence and a premise sentence. Datumo\u2019s crowd-workers constructed numerous true\/false\/neutral sentences based on the premise and classified the sentences made by other workers back as true\/false\/neutral.<\/p>\n<p>&nbsp;<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-15226\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-8.53.48-PM.png\" alt=\"\" width=\"620\" height=\"276\" \/><\/p>\n<p>&nbsp;<\/p>\n<p>KLUE-NLI, based on Datumo&#8217;s data, performed higher accuracy compared to SNLI(Stanford Natural Language Inference) and MNLI(Multi-Genre Natural Language Inference).<\/p>\n<table class=\" alignleft\" style=\"height: auto; width: 100%; border-collapse: collapse; border-style: none;\">\n<thead>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #cccccc; border-style: none; width: 32.518%; height: auto;\"><strong>Statistics<\/strong><\/td>\n<td style=\"border-color: #cccccc; border-style: none; width: 30.6475%; height: auto;\"><strong>SNLI<\/strong><\/td>\n<td style=\"border-color: #cccccc; border-style: none; width: 205.756%; height: auto;\"><strong>MNLI<\/strong><\/td>\n<td style=\"border-color: #cccccc; border-style: none; width: 56.0792%; height: auto;\"><strong>KLUE-NLI<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 32.518%; height: auto;\">Unanimous<br \/>\nGold Label<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 30.6475%; height: auto;\">58.30%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 205.756%; height: auto;\">58.20%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>76.29%<\/strong><\/span><\/td>\n<\/tr>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; width: 32.518%; height: auto;\">Individual Label<br \/>\n= Gold Label<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 30.6475%; height: auto;\">89.00%<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 205.756%; height: auto;\">88.70%<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>92.63%<\/strong><\/span><\/td>\n<\/tr>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; width: 32.518%; height: auto;\">Individual Label<br \/>\n= Author\u2019s Label<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 30.6475%; height: auto;\">85.80%<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 205.756%; height: auto;\">85.20%<\/td>\n<td style=\"border-color: #ededed; border-style: none; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>90.92%<\/strong><\/span><\/td>\n<\/tr>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 32.518%; height: auto;\">Gold Label<br \/>\n= Author\u2019s Label<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 30.6475%; height: auto;\">91.020%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 205.756%; height: auto;\">92.60%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>96.76%<\/strong><\/span><\/td>\n<\/tr>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 32.518%; height: auto;\">Gold Label<br \/>\n\u2260 Author\u2019s Label<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 30.6475%; height: auto;\">6.80%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 205.756%; height: auto;\">5.60%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>2.71%<\/strong><\/span><\/td>\n<\/tr>\n<tr style=\"height: auto;\">\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 32.518%; height: auto;\">No Gold Label<br \/>\n(No 3 Labels Match)<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 30.6475%; height: auto;\">2.00%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 205.756%; height: auto;\">1.80%<\/td>\n<td style=\"border-color: #ededed; border-style: none; background-color: #f5f5f5; width: 56.0792%; height: auto;\"><span style=\"font-size: 14pt;\"><strong>0.53%<\/strong><\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>We call the label intended by the original annotator in writing the hypothesis &#8220;author\u2019s label.&#8221; Consensus among three out of five annotators is &#8220;gold label.&#8221;_Source: KLUE Paper[\/vc_column_text]<div id=\"el1650446598380-8981bcec-7b26\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h3 class=\"text-body-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Machine Reading Comprehension, MRC<\/h3><\/div><\/div><\/div><div id=\"el1650437452756-494cad23-44db\" class=\"w-100 d-block \"><\/div>[vc_column_text css=&#8221;.vc_custom_1653303119151{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;]KLUE\u2019s Machine reading comprehension (MRC) is a task designed to evaluate models\u2019 abilities to read a given text passage and then answer a question about the passage. In order to enable the model to \u201ccomprehend\u201d the given text, a number of crowd-workers formulated a pair of sentences(question and answer) and the sentences went through validation.<\/p>\n<p>&nbsp;<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-15229\" src=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-9.09.58-PM-1024x571.png\" alt=\"\" width=\"640\" height=\"357\" \/><\/p>\n<p>&nbsp;<\/p>\n<p>Text<\/p>\n<ul>\n<li>According to The Hydrometeorological Centre of Russia, Moscow, known for its cold weather, warmed up to 7 degrees Celsius in the afternoon of22nd(local time).<\/li>\n<li>Goldman Sachs, an American investment bank and financial services company, have reported on 22nd that the consumption of natural gas and oil for heating is decreasing and that oversupply of oil and hesitance in global economic growth will even lower the cost per unit of energy.<\/li>\n<\/ul>\n<p>Question<\/p>\n<ul>\n<li>What was the highest temperature in Moscow on the day Goldman Sachs published their report?<\/li>\n<\/ul>\n<p>Answer<\/p>\n<ul>\n<li>7 degrees Celsius<\/li>\n<\/ul>\n<p>An AI model trained with KLUE\u2019s MRC dataset should be able to \u201ccomprehend\u201d the context and be able to infer an answer for a question. As such, AI is only as smart as the data that feeds it, and that is why achieving high quality data is so crucial.[\/vc_column_text]<div id=\"el1650850456269-171f3fc9-aebd\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1650851360230-b30a5134-5e08\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-left \"><div><div class=\"slide-in-container\"><h2 class=\"text-heading-default font-weight-bold heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"\" data-anim-delay=\"0\">Project Retrospective<\/h2><\/div><\/div><\/div><div id=\"el1650294934292-07172d41-57be\" class=\"w-100 d-block \"><\/div>[vc_row_inner css=&#8221;.vc_custom_1653305210643{padding-right: 16px !important;padding-left: 16px !important;}&#8221;][vc_column_inner width=&#8221;1\/6&#8243;]<div class=\"pix-story d-inline-block \"  ><div class=\"story-img pix-bg-attachment-scroll pix-bg-custom shadow scale-inverse-sm pix-no-bg text-left d-inline-block\" ><\/div><\/div>[\/vc_column_inner][vc_column_inner width=&#8221;5\/6&#8243;]<div id=\"el1650514662141-d782b60a-8f61\" class=\"w-100 d-block \"><\/div>[vc_column_text]<strong>Dongyeon Kim | Project Manager<\/strong>[\/vc_column_text][vc_column_text]<\/p>\n<p style=\"text-align: left;\">\u201cAmong the eight KLUE project tasks, Datumo collected and annotated four: TC, STS, NLI, MRC. I was in charge of the MRC task and there were three types within the task.<\/p>\n<p>[\/vc_column_text]<div class=\"pix-content-box card      custom-responsive-54573076   rounded-lg shadow-sm bg- w-100  \"   ><div class=\"\" style=\"z-index:30;position:relative;\"><div id=\"el1651553796120-92effbc4-0b96\" class=\"mb-3 text-left \"><h5 class=\"mb-32 pix-sliding-headline-2 animate-in font-weight-bold secondary-font \" data-anim-type=\"pix-sliding-text\" pix-anim-delay=\"500\" data-class=\"secondary-font text-body-default\" style=\"\"><span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 0ms;\">1.&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 150ms;\">Paraphrasing&#32;<\/span><\/span> <\/h5><\/div><div id=\"el1651553821837-88c501c3-2504\" class=\"mb-3 text-left \"><h5 class=\"mb-32 pix-sliding-headline-2 animate-in font-weight-bold secondary-font \" data-anim-type=\"pix-sliding-text\" pix-anim-delay=\"500\" data-class=\"secondary-font text-body-default\" style=\"\"><span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 0ms;\">2.&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 150ms;\">Multi-sentence&#32;<\/span><\/span> <\/h5><\/div><div id=\"el1651553824065-5afd33ba-3b90\" class=\"mb-3 text-left \"><h5 class=\"mb-32 pix-sliding-headline-2 animate-in font-weight-bold secondary-font \" data-anim-type=\"pix-sliding-text\" pix-anim-delay=\"500\" data-class=\"secondary-font text-body-default\" style=\"\"><span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 0ms;\">3.&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-body-default\" style=\"transition-delay: 150ms;\">Non-answerable&#32;<\/span><\/span> <\/h5><\/div><\/div><\/div>[vc_column_text]<\/p>\n<p style=\"text-align: left;\">There were some challenges such as setting a standard for acceptable paraphrasing and combining sentences to avoid situations where there could be multiple correct answers. Some text snippets were Korean translations from English, which made some sentences sound unnatural or awkward, whereas some had too many jargons. However, after continuous communication, we were able to come up with a method to maintain the consistency and quality of the final dataset.\u201d<\/p>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner]<div id=\"el1650294934292-37bdaa6b-ef9a\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][vc_section full_width=&#8221;stretch_row&#8221; pix_over_visibility=&#8221;&#8221; pix_visibility=&#8221;1&#8243; css=&#8221;.vc_custom_1647419266512{border-top-width: 1px !important;padding-top: 80px !important;padding-bottom: 80px !important;background-color: #ffffff !important;border-top-color: rgba(0,0,0,0.11) !important;border-top-style: solid !important;}&#8221;][vc_row full_width=&#8221;stretch_row_content&#8221; content_placement=&#8221;middle&#8221; pix_particles_check=&#8221;&#8221;][vc_column content_align=&#8221;text-center&#8221; offset=&#8221;vc_col-md-offset-2 vc_col-md-8&#8243;]<h6 class=\"pix-badge-element d-inline-block mr-1 animate-in\" data-anim-type=\"fade-in-up\" data-anim-delay=\"300\"><span class=\"badge font-weight-bold secondary-font bg-gray-2  \" style=\" \"><span class=\"text-body-default\" style=\"\">Open Datasets for Data-Centric AI<\/span><\/span><\/h6><div id=\"el1647418693885-021e86c3-1643\" class=\"mb-3 text-center \"><h2 class=\"mb-32 pix-sliding-headline-2 animate-in font-weight-bold secondary-font \" data-anim-type=\"pix-sliding-text\" pix-anim-delay=\"500\" data-class=\"secondary-font text-black\" style=\"\"><span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 0ms;\">The&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 150ms;\">above&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 300ms;\">datasets&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 450ms;\">can&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 600ms;\">be&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 750ms;\">downloaded&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 900ms;\">for&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 1050ms;\">free&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 1200ms;\">through&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 1350ms;\">DATUMO&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 1500ms;\">\u2018OPEN&#32;<\/span><\/span> <span class=\"slide-in-container \"><span class=\"pix-sliding-item secondary-font text-black\" style=\"transition-delay: 1650ms;\">DATASETS'.&#32;<\/span><\/span> <\/h2><\/div>[\/vc_column][\/vc_row][vc_row full_width=&#8221;stretch_row&#8221; pix_particles_check=&#8221;&#8221;][vc_column css=&#8221;.vc_custom_1591419244990{margin-bottom: -80px !important;}&#8221; offset=&#8221;vc_col-md-offset-1 vc_col-md-10&#8243;]<a href=\"https:\/\/open.selectstar.ai\/\" target=\"_blank\" title=\"\"><div class=\"pix-fancy-mockup\"><div class=\"pix-fancy-content\"><img decoding=\"async\" class=\"\" src=\"\" alt=\"\"><\/div>\n                    <img decoding=\"async\" class=\"pix-fancy-device-img\" src=\"https:\/\/dev.selectstar.ai\/ko\/wp-content\/plugins\/pixfort-core\/functions\/images\/ipad.png\" alt=\"\"><\/div><\/a>[\/vc_column][\/vc_row][vc_row full_width=&#8221;stretch_row_content&#8221; content_placement=&#8221;middle&#8221; pix_particles_check=&#8221;&#8221;][vc_column content_align=&#8221;text-center&#8221; offset=&#8221;vc_col-md-offset-2 vc_col-md-8&#8243;]<div id=\"el1647419546778-06abc118-76ab\" class=\"w-100 d-block \"><\/div><div  class=\"pix-heading-el text-center \"><div><div class=\"slide-in-container\"><h4 class=\"text-gray-7 font-weight-bold secondary-font animate-in heading-text el-title_custom_color mb-12\" style=\"\" data-anim-type=\"fade-in-left\" data-anim-delay=\"400\">DATUMO would like to support the AI industry by sharing.<\/h4><\/div><\/div><\/div><a  href=\"https:\/\/open.datumo.com\/en\/\" class=\"btn mb-2     shadow-sm text-white btn-black d-inline-block      animate-in btn-lg\" target=\"_blank\" rel=\"noopener\"  data-anim-type=\"fade-in-left\" data-anim-delay=\"1000\" ><span class=\"font-weight-bold \" >Learn More About DATUMO Open Datasets<\/span><\/a><div id=\"el1647419426201-0817d40b-b576\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row][\/vc_section][vc_row pix_particles_check=&#8221;&#8221;][vc_column width=&#8221;1\/2&#8243;]<div id=\"el1647418606245-ee6aac62-9c1b\" class=\"w-100 d-block \"><\/div><div class=\"d-inline-block \" ><div class=\"pix-img-el    text-left d-inline-block  w-100 rounded-0\"  ><\/div><\/div><div id=\"el1647416874497-981daf55-c2d2\" class=\"w-100 d-block \"><\/div>[\/vc_column][vc_column width=&#8221;1\/2&#8243;]<div id=\"el1647418614985-7fc6233e-60a5\" class=\"w-100 d-block \"><\/div>[vc_column_text]<\/p>\n<p class=\"indexstyle__BottomTitle-sc-1ntcad2-8 eXeKgC\">CC BY-SA<\/p>\n<p class=\"indexstyle__BottomDescription-sc-1ntcad2-9 bKpFwB\">Reusers are allowed to distribute, remix, adapt, and build upon the material in any medium or format, even commercially, so long as attribution is given to the creator. If you remix, adapt, or build upon the material, you must license the modified material under identical terms.<\/p>\n<p class=\"indexstyle__BottomDescription-sc-1ntcad2-9 bKpFwB\">https:\/\/creativecommons.org\/licenses\/by-sa\/3.0\/deed.en<\/p>\n<p>[\/vc_column_text][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1646794934167-c0c94dd3-ea74\" class=\"w-100 d-block \"><\/div><div class=\" mb-3 mb-md-0 \"  ><div class=\"card w-100 h-100 bg-white   pix-hover-item rounded-10 position-relative overflow-hidden2 text-white tilt fancy_card\" ><div class=\"card-img-overlay overflow-visible d-inline-block w-100 pix-img-overlay pix-p-30 d-flex align-items-end text-left\"><div class=\"w-100 \"><h3 class=\"card-title  text-black font-weight-bold mb-0 animate-in\" style=\"\">See what we can do for you.<\/h3><p class=\"card-text pix-pt-10 text-black \" style=\"\">Build smarter AI with us.<\/p><div class=\"card-btn-div mt-4 d-inline-block w-100\"><a  href=\"https:\/\/datumo.com\" class=\"btn mb-2     text-white btn-black d-inline-block      btn-md\" target=\"_blank\" rel=\"noopener\"    ><span class=\"font-weight-bold \" >Learn More<\/span><\/a><\/div><\/div><\/div><\/div><\/div>[\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column]<div id=\"el1646799961152-e3ee06c0-4e82\" class=\"w-100 d-block \"><\/div>[\/vc_column][\/vc_row]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>[vc_row pix_particles_check=&#8221;&#8221;][vc_column][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_row_inner][vc_column_inner width=&#8221;1\/5&#8243;][\/vc_column_inner][vc_column_inner width=&#8221;4\/5&#8243;][vc_column_text] Korean Language Understanding\u00a0Evaluation Benchmark [\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_column_text css=&#8221;.vc_custom_1653300469410{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;] The KLUE(Korean Language Understanding\u00a0Evaluation Benchmark) paper, written using Datumo\u2019s datasets, has been accepted in NeurIPS 2021, a world-renowned AI conference. Natural&#8230;<\/p>\n","protected":false},"author":1,"featured_media":14429,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_uag_custom_page_level_css":"","footnotes":""},"categories":[44],"tags":[],"table_tags":[],"class_list":["post-15175","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-usecases_success"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.0 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dev.selectstar.ai\/ko\/klue\/\" \/>\n<meta property=\"og:locale\" content=\"ko_KR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\" \/>\n<meta property=\"og:description\" content=\"[vc_row pix_particles_check=&#8221;&#8221;][vc_column][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_row_inner][vc_column_inner width=&#8221;1\/5&#8243;][\/vc_column_inner][vc_column_inner width=&#8221;4\/5&#8243;][vc_column_text] Korean Language Understanding\u00a0Evaluation Benchmark [\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_column_text css=&#8221;.vc_custom_1653300469410{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;] The KLUE(Korean Language Understanding\u00a0Evaluation Benchmark) paper, written using Datumo\u2019s datasets, has been accepted in NeurIPS 2021, a world-renowned AI conference. Natural...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dev.selectstar.ai\/ko\/klue\/\" \/>\n<meta property=\"og:site_name\" content=\"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\" \/>\n<meta property=\"article:published_time\" content=\"2022-04-20T03:53:28+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-2.07.40-PM-1024x457.png\" \/>\n<meta name=\"author\" content=\"selectstar\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"selectstar\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"13\ubd84\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/\"},\"author\":{\"name\":\"selectstar\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/person\/22d8a968b9bae37589037ecc7d84dfba\"},\"headline\":\"Open Datasets &#8211; KLUE\",\"datePublished\":\"2022-04-20T03:53:28+00:00\",\"dateModified\":\"2022-04-20T03:53:28+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/\"},\"wordCount\":2644,\"publisher\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#organization\"},\"articleSection\":[\"use cases\"],\"inLanguage\":\"ko-KR\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/\",\"url\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/\",\"name\":\"Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\",\"isPartOf\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#website\"},\"datePublished\":\"2022-04-20T03:53:28+00:00\",\"dateModified\":\"2022-04-20T03:53:28+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/#breadcrumb\"},\"inLanguage\":\"ko-KR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/dev.selectstar.ai\/ko\/klue\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/klue\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/dev.selectstar.ai\/ko\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Open Datasets &#8211; KLUE\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#website\",\"url\":\"https:\/\/dev.selectstar.ai\/ko\/\",\"name\":\"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\",\"description\":\"Your AI Data Standard\",\"publisher\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/dev.selectstar.ai\/ko\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"ko-KR\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#organization\",\"name\":\"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\",\"url\":\"https:\/\/dev.selectstar.ai\/ko\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/dev.selectstar.ai\/ko\/wp-content\/uploads\/2024\/08\/cropped-sb-1.webp\",\"contentUrl\":\"https:\/\/dev.selectstar.ai\/ko\/wp-content\/uploads\/2024\/08\/cropped-sb-1.webp\",\"width\":1276,\"height\":167,\"caption\":\"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0\"},\"image\":{\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/logo\/image\/\"}},{\"@type\":\"Person\",\"@id\":\"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/person\/22d8a968b9bae37589037ecc7d84dfba\",\"name\":\"selectstar\",\"url\":\"https:\/\/dev.selectstar.ai\/ko\/author\/selectstar\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dev.selectstar.ai\/ko\/klue\/","og_locale":"ko_KR","og_type":"article","og_title":"Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","og_description":"[vc_row pix_particles_check=&#8221;&#8221;][vc_column][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_row_inner][vc_column_inner width=&#8221;1\/5&#8243;][\/vc_column_inner][vc_column_inner width=&#8221;4\/5&#8243;][vc_column_text] Korean Language Understanding\u00a0Evaluation Benchmark [\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_column_text css=&#8221;.vc_custom_1653300469410{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;] The KLUE(Korean Language Understanding\u00a0Evaluation Benchmark) paper, written using Datumo\u2019s datasets, has been accepted in NeurIPS 2021, a world-renowned AI conference. Natural...","og_url":"https:\/\/dev.selectstar.ai\/ko\/klue\/","og_site_name":"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","article_published_time":"2022-04-20T03:53:28+00:00","og_image":[{"url":"https:\/\/blog.datumo.com\/en\/wp-content\/uploads\/2022\/04\/Screen-Shot-2022-04-13-at-2.07.40-PM-1024x457.png"}],"author":"selectstar","twitter_card":"summary_large_image","twitter_misc":{"Written by":"selectstar","Est. reading time":"13\ubd84"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/#article","isPartOf":{"@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/"},"author":{"name":"selectstar","@id":"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/person\/22d8a968b9bae37589037ecc7d84dfba"},"headline":"Open Datasets &#8211; KLUE","datePublished":"2022-04-20T03:53:28+00:00","dateModified":"2022-04-20T03:53:28+00:00","mainEntityOfPage":{"@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/"},"wordCount":2644,"publisher":{"@id":"https:\/\/dev.selectstar.ai\/ko\/#organization"},"articleSection":["use cases"],"inLanguage":"ko-KR"},{"@type":"WebPage","@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/","url":"https:\/\/dev.selectstar.ai\/ko\/klue\/","name":"Open Datasets - KLUE - \ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","isPartOf":{"@id":"https:\/\/dev.selectstar.ai\/ko\/#website"},"datePublished":"2022-04-20T03:53:28+00:00","dateModified":"2022-04-20T03:53:28+00:00","breadcrumb":{"@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/#breadcrumb"},"inLanguage":"ko-KR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dev.selectstar.ai\/ko\/klue\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/dev.selectstar.ai\/ko\/klue\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dev.selectstar.ai\/ko\/"},{"@type":"ListItem","position":2,"name":"Open Datasets &#8211; KLUE"}]},{"@type":"WebSite","@id":"https:\/\/dev.selectstar.ai\/ko\/#website","url":"https:\/\/dev.selectstar.ai\/ko\/","name":"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","description":"Your AI Data Standard","publisher":{"@id":"https:\/\/dev.selectstar.ai\/ko\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dev.selectstar.ai\/ko\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"ko-KR"},{"@type":"Organization","@id":"https:\/\/dev.selectstar.ai\/ko\/#organization","name":"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0","url":"https:\/\/dev.selectstar.ai\/ko\/","logo":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/logo\/image\/","url":"https:\/\/dev.selectstar.ai\/ko\/wp-content\/uploads\/2024\/08\/cropped-sb-1.webp","contentUrl":"https:\/\/dev.selectstar.ai\/ko\/wp-content\/uploads\/2024\/08\/cropped-sb-1.webp","width":1276,"height":167,"caption":"\ub370\uc774\ud130\ub294 \uc140\ub809\ud2b8\uc2a4\ud0c0"},"image":{"@id":"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"https:\/\/dev.selectstar.ai\/ko\/#\/schema\/person\/22d8a968b9bae37589037ecc7d84dfba","name":"selectstar","url":"https:\/\/dev.selectstar.ai\/ko\/author\/selectstar\/"}]}},"uagb_featured_image_src":{"full":false,"thumbnail":false,"medium":false,"medium_large":false,"large":false,"1536x1536":false,"2048x2048":false,"pix-blog-small":false,"pix-portfolio-small":false,"pix-big":false,"pix-square-sm":false,"pix-woocommerce-xs":false,"pix-woocommerce-md":false,"pix-xxl":false},"uagb_author_info":{"display_name":"selectstar","author_link":"https:\/\/dev.selectstar.ai\/ko\/author\/selectstar\/"},"uagb_comment_info":0,"uagb_excerpt":"[vc_row pix_particles_check=&#8221;&#8221;][vc_column][\/vc_column][\/vc_row][vc_row pix_particles_check=&#8221;&#8221;][vc_column][vc_row_inner][vc_column_inner width=&#8221;1\/5&#8243;][\/vc_column_inner][vc_column_inner width=&#8221;4\/5&#8243;][vc_column_text] Korean Language Understanding\u00a0Evaluation Benchmark [\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_column_text css=&#8221;.vc_custom_1653300469410{padding-top: 40px !important;padding-right: 20px !important;padding-bottom: px !important;padding-left: 20px !important;}&#8221;] The KLUE(Korean Language Understanding\u00a0Evaluation Benchmark) paper, written using Datumo\u2019s datasets, has been accepted in NeurIPS 2021, a world-renowned AI conference. Natural...","_links":{"self":[{"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/posts\/15175","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/comments?post=15175"}],"version-history":[{"count":0,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/posts\/15175\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/"}],"wp:attachment":[{"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/media?parent=15175"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/categories?post=15175"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/tags?post=15175"},{"taxonomy":"table_tags","embeddable":true,"href":"https:\/\/dev.selectstar.ai\/ko\/wp-json\/wp\/v2\/table_tags?post=15175"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}